Inflation is falling without a recession. Can it last?

Consumer price inflation is slowing while the US economy still adds more than 200,000 jobs a month, and the optimists have found their gospel. Believers in 'immaculate disinflation' argue that prices can be brought back under control without a labour-market recession to pay for it. The story is appealing. It also skips over how economic cycles usually adjust.

The case the optimists make

The evidence for their view is real. Headline consumer price inflation has fallen from 9.1 per cent in June last year to 4 per cent in May, and economists expect the June figure, due this week, to drop further as last summer's energy spike falls out of the comparison. Over the same period employers kept hiring. The June jobs report showed 209,000 new positions and an unemployment rate of 3.6 per cent, close to the lowest in half a century.

Job vacancies have also fallen a long way from their peak without pushing up unemployment. Two Fed economists, one of them the governor Christopher Waller, argued last year that the labour market could cool mainly through fewer openings rather than more layoffs. Olivier Blanchard and Larry Summers took the opposite side, warning that history offered few examples of vacancies falling that far without joblessness rising. So far the data have been kinder to Waller.

Supply chains have done their bit

The first stage of disinflation came from supply chains unjamming after the pandemic and from retailers clearing bloated stocks of goods. That was a gift from the supply side, and it cooled prices without anyone having to crush demand. The gift has now largely been spent. Getting inflation from three per cent to two per cent means taming services prices, which takes either a jump in productivity or a looser labour market. Productivity miracles are rare.

The composition of recent inflation makes the point. Energy prices have fallen outright from their peak, and many goods prices are flat or lower, as shipping costs have collapsed and car production has recovered. Core inflation, which excludes food and energy, was still 5.3 per cent in May. The difference between the two measures is almost entirely the supply-side gift, and it cannot be repeated because energy and freight costs cannot fall from where they already are by the same amount again.

What services need

Services prices depend chiefly on wages, and wage growth depends on how many workers employers are chasing. With unemployment near its lowest levels in decades, companies still compete for staff, and pay rises remain faster than is consistent with 2 per cent inflation unless productivity grows unusually quickly. Measured productivity has in fact been weak since the pandemic, which leaves more of each wage increase to be passed on in prices.

The Federal Reserve has drawn its own conclusion. It paused in June after ten consecutive increases but signalled two more before the end of the year, and markets expect the first of those at the end of July. Officials would not be planning further tightening if they believed the job market could stay this tight while inflation kept falling.

The route from higher rates to a softer labour market runs mainly through credit. Banks, shaken by the failures of Silicon Valley Bank and its peers in the spring, have tightened lending standards, according to the Fed's own surveys of loan officers. Firms that cannot borrow cheaply postpone expansion, then slow hiring, and only later cut staff. That sequence usually takes a year or more to work through, which is why the full effect of last year's rate rises has probably not yet reached the jobs figures. The optimists read the resilient payroll numbers as proof the economy can absorb high rates. They may simply be early readings from a process still under way.

The rare soft landing

Soft landings are not impossible. In 1994 the Fed doubled its policy rate from 3 to 6 per cent within a year, inflation stayed contained and the economy kept growing for the rest of the decade. That episode is the model optimists cite. The big difference is that the Fed in 1994 was pre-empting inflation that had not yet arrived, whereas today it is trying to reverse inflation that has already become embedded in wages and prices.

The other hopeful precedent is the late 1990s, when a surge in productivity linked to information technology let the economy run with low unemployment and low inflation at the same time. It is possible that new technology, or the reallocation of workers after the pandemic, will deliver something similar. Economists have predicted such booms far more often than they have seen them.

Who gains if it works, and who pays if not

If the optimists are right, the winners are obvious: workers keep their jobs, companies keep their revenues, and equity investors who bought on the hope of a soft landing are vindicated. If they are wrong, the adjustment will fall most heavily on lower-paid service workers, whose jobs are the first to go when demand softens, and on small firms that borrowed on floating rates. Bond investors, by contrast, would benefit from a slowdown that forces the Fed to cut.

The strongest counterargument is that part of the labour market's heat reflects temporary distortions. Retirements during the pandemic, lower immigration and mismatches between workers and jobs pushed vacancies to unusual highs, and as those fade, wage pressure could ease without layoffs. There is something in this. Immigration has recovered, and the pool of available workers has grown. But it is a claim that the labour market can loosen painlessly, not evidence that services inflation can fall while it stays tight.

The figures to watch are monthly payrolls and unemployment, the Fed's preferred measure of core services inflation, and wage data such as the employment cost index. The painless part of disinflation is ending. Getting the rest of the way to target without a recession will take more luck than central banks usually enjoy, and the labour market will probably have to soften first. The likeliest sign will be unemployment drifting up from today's 3.6 per cent over the coming year, with lower-paid service workers bearing the brunt.